1. Introduction
In recent years, timber structures have experienced considerable development throughout the world, owing to their ‘green’ and low-carbon properties. Research on timber structures has become a significant focus worldwide, especially in North America and Europe, and even in China [
1].
As is well-known, joints or connections are important aspects of timber structures [
2,
3]. Among the various types of joints, the glued-in rod joint shows excellent structural performances, with a pronounced bearing capacity and considerable slip stiffness, and it can be applied to high-rise and large-span timber structures [
4,
5,
6]. It has been demonstrated that the glued-in rod/rebar joints showed superior feasibility and mechanical properties while they was applied to prefabricated beam-to-column connection [
7,
8].
Many experimental and theoretical studies on glued-in rebar timber joints have been carried out. Broughton and Hutchinson [
9] investigated the influence of adhesive types on the behaviour of bonded steel rod joints and found that epoxy resin shows the best bearing capacity. Chans et al. [
10] investigated the influence of geometric and material characteristics on the performance of glued-in rod joints. The joint performance showed evident differences when the joint was parallel [
11] and perpendicular [
12] to the grain directions. Azinović et al. [
13] evaluated the pull-out strength of glued-in rods in cross-laminated timber and obtained the pull-out failure modes. Vallée and Adam [
14] significantly shortened the curing time of glued-in rods in timber to minutes by using Curie particles. Ling et al. [
15] proposed an empirical model for predicting the bond-slip behaviour of glued-in rebar joints, and also conducted pull-out tests by considering timber species, rod types, adhesive types, bonded length, glue-line thickness, and angle with respect to the grain [
16]. Fragiacomo and Batchelar [
17,
18] presented a design method for evaluating joint strength under the combined action of moment and axial forces, and also evaluated the long-term performance of glued-in steel rods. Xu et al. [
19] researched the effects of manufacturing defects on the pull-out behaviour of glued-in rods, and concluded that the positioning defects of the rods can be negligible. Grunwald et al. [
20,
21] investigated the glued-in rod joints which were applied to engineered hardwood products, and found that the shear strength and transverse (tensile) strength of wood significantly influence the bearing capacities of joints, while longitudinal tensile strength proved to be less significant. Similarly, He & Xiao [
22] also demonstrated that the glued-in rod joints can be applied to glued laminated bamboo structures, and showed significant mechanical properties in bearing capacity and slip stiffness.
Hence, research on glued-in rod joints at room temperature is comprehensive and detailed. However, considering the importance of the structural behaviour of timber joints under fire conditions, it is extremely urgent to investigate their fire resistance [
23]. Some studies have researched the fire resistance of some traditional joints, such as tooth plate joints [
24], bolted joints [
25,
26,
27,
28,
29], and dowel joints [
28,
29,
30,
31]. Liu et al. [
32] investigated the bonding performance of glulam with different adhesives at elevated temperature, and found bonding performances of structural adhesives deteriorated linearly with increasing temperature.
To improve the fireproof ratings of timber members, surface impregnation treatment [
33] and encapsulation coating system are common methods, which were suggested in available design methods [
34,
35]. As to the surface impregnation treatment, two-hour fireproof performance can be obtained for cross laminated timber while it was covered with fire-retardant impregnated wood [
36]. In addition, Yue et al. [
37] improved mechanical properties and combustion performance of Chinese fir by using impregnation and densification methods. The encapsulation coating system includes the fire-retardant coating, gypsum boards, and suspended membrane-type encapsulation [
38]. Popescu and Pfriem [
39] summarised the existing investigations about treatments and modification to improve the fire resistant capacity of wood and wood based products. Applying concrete slabs to the top of timber joists was an effective method to improve the fireproof performance of timber members [
40]. In addition, for the fire design of timber structures, the reduction section method by considering the char rating of timber was adopted in available standards [
41,
42].
However, only a few research studies have focused on the fire resistance of glued-in rod/rebar joints. Lartigau et al. [
43] studied the influence of temperature on the mechanical properties of glued-in rebar timber joints with epoxy resins. The study found that the critical temperature was 58 °C, and the pull-out strength evidently decreased when the temperature surpassed 58 °C. Harris [
44] performed pull-out experiments for glued-in rebar joints in laminated veneer lumber (LVL) by considering a hybrid adhesive composed of urethane methacrylate and cement, aiming to improve the fire resistance of epoxy resin. However, the structural performance of the proposed modified adhesives in room conditions was not very satisfactory. Park et al. [
45] carried out fire experiments on glued-in rebar joints in which two epoxy resins and a composite adhesive (HY150) were adopted, and the experimental results showed that the residual strength of epoxy at 100 °C was reduced by 30% from its cold strength, whereas the composite adhesive showed a smaller strength loss. In conclusion, the fire-resistant adhesive of glued-in rod joints proposed in the existing researches showed the obvious disadvantage in bearing capacity at room temperature.
According to the available literatures, the existing researches on the fire-resistance of glued-in rod joints has mainly focused on those with a traditional and commercial epoxy resin. Only a few heat resistant modified adhesives have been proposed [
44,
45]. For this reason, this study proposed a type of heat resistant modified waterborne epoxy resin by introduction of an inorganic refractory filler aiming to improve the fire resistant performance of adhesive used in glued-in rod joints. Firstly, preliminary pull-out tests of the modified epoxy resin were performed at room temperature, and the test results showed satisfactory mechanical behaviours compared with the commercial epoxy resin. Subsequently, fire tests were conducted under insurance service office- (ISO-) fire conditions. Testing results showed that the glued-in rod joints using heat resistant modified epoxy resin showed evident improvement in fire-resistant time in fire tests, without significant decline in mechanical performance at room temperature.
4. Bonding Behaviour at Room Temperature
To investigate the bonding strength of the modified epoxy adhesive, pull-out tests were conducted at room temperature. The modified epoxy adhesive and commercial epoxy resin were adopted separately by the pull-pull specimens. Three samples were prepared for each type of adhesive. For all specimens, the anchorage length was 240 mm, whereas the thickness of the glue line was 3 mm. The anchorage length and glue line thickness were consistent with that of the specimens in the fire testing. The results of the pull-out tests at room temperature are depicted in
Figure 2. Of particular note is that the values of slip tested in pull-out tests at room temperature were the displacement of the actuator aiming to being consistent with the measurement scheme in fire tests, instead of the interface relative slip between rod and timber which were commonly adopted in available pull-out tests at room conditions [
15].
The average bearing capacity for the specimens with traditional commercial epoxy resin (E240-3) is 106.8 kN, whereas that for specimens with the modified epoxy adhesive (M240-3) is 114.2 kN. In addition, two groups of slip curves showed similar slopes denoting that the slip stiffness for the glued-in rod joint specimens with the two types of adhesives adopted has no obvious difference.
Figure 3 shows the failure photographs of tested glued-in rod joints. The specimens with heat resistant modified epoxy adhesive and commercial epoxy resin showed a same failure mode that was the shear failure of timber. In addition, the commercial epoxy resin showed the dark green as shown in
Figure 3b, whereas the modified epoxy adhesive was almost black as shown in
Figure 3a as the result of the application of cement and carbon fibre powders (see
Table 4). Therefore, the modified epoxy adhesive shows similar mechanical performances compared with commercial epoxy resin at room temperature. Thus, the modified epoxy adhesive used in this study can be applied to glued-in rebar/rod timber joints in practical engineering in terms of mechanical behaviour at room temperatures.
8. Conclusions
In this work, an innovative modified fireproof epoxy resin adhesive was proposed, based on introducing some inorganic additives. Through tests both at room temperature and in fire, the effects on the fire resistance of joints from several influential factors were analysed in detail. The main conclusions of this study are summarised as follows:
As far as mechanical behaviours are concerned, there is little difference between a glued-in rod timber joint with the heat-resistant modified epoxy resin and that with the commercial epoxy resin at room temperature.
The fire resistance period of the specimens with the modified epoxy resin can be improved by 73.0% when compared with the specimens with the commercial epoxy resin.
The fire resistance time declined to 57 min from 70 min, while the loading level in fire tests increased to 50% from 20%. The loading levels sustained by the stress component should be emphasized.
When the specimens were protected by a white coating, the fire resistance period increased by 54.1% and 20.3% for the traditional and modified epoxy resin cases, respectively.
As the joint edge distance changed from 5d to 7.5d, the failure time increased from 64 min to 83 min, an improvement of 29.7%.
Under the same load level and sectional size, the specimens in group ZF160-3, which adopted the modified epoxy adhesive and white fire coating, showed the best fire resistance performance.
In conclusion, the modified epoxy adhesive showed the significant improvement in fire resistance compared with the ordinary commercial epoxy resin, without any decline in mechanical performance at room temperatures. Therefore, the modified epoxy adhesive proposed in this article exhibits superior application prospect in glued-in rod timber joints thanks to the advantage of heat resistant.
Further investigations will focus on the theoretical analysis about determining the temperature tendency of adhesive based on the charring rate and thermal conductivity of timber, as well as the stiffness reduction factor of glued-in rod joints with different adhesives in fires. Thus, the relationship among fire time, adhesive temperature, and joint stiffness can be established to guide the fire design of glued-in rods joints of timber structures.